• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.05a
• /
• pp.515-518
• /
• 2019

Semantic segmentation has remained as a challenging problem in the field of computer vision. Given the immense power of Convolution Neural Network (CNN) models, many complex problems have been solved in computer vision. Semantic segmentation is the challenge of classifying several pixels of an image into one category. With the help of convolution neural networks, we have witnessed prolific results over the time. We propose a convolutional neural network model which uses Fully CNN with deconvolutional pixel layers. The goal is to create a hierarchy of features while the fully convolutional model does the primary learning and later deconvolutional model visually segments the target image. The proposed approach creates a direct link among the several adjacent pixels in the resulting feature maps. It also preserves the spatial features such as corners and edges in images and hence adding more accuracy to the resulting outputs. We test our algorithm on Karlsruhe Institute of Technology and Toyota Technologies Institute (KITTI) street view data set. Our method achieves an mIoU accuracy of 92.04 %.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2015.07a
• /
• pp.540-541
• /
• 2015

본 논문은 카메라 촬영 시 손 떨림 등에 의해 생기는 불균일한 블러 현상에 대해 영상 IMU 센서와 노출시간이 긴 영상과 짧은 영상 쌍을 이용한 디블러 알고리즘을 제안한다. 먼저 IMU 센서로부터 얻어진 gyro 데이터로 초기 커널을 추정한다. 그런 다음 노출시간이 다른 영상 쌍에 Lucas-Kanade 기법을 적용하여 상기 추정된 초기 커널을 개선한다. 이렇게 구해진 커널에 기반하여 residual deconvolution 을 수행 디블러 영상을 생성한다. 실험 결과로부터 제안 알고리즘이 기존 기법에 비해 우수한 성능을 보임을 알 수 있다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2015.11a
• /
• pp.114-117
• /
• 2015

본 논문에서는 Non-blind 디컨벌루션 알고리듬 중 하나인 Richardson-Lucy(RL) 디컨벌루션을 영상 분할을 통해 성능을 향상시킨 알고리듬을 제안한다. RL 디컨벌루션은 영상의 크기가 커질수록 연산 양이 크게 증가한다. 따라서 크기가 큰 영상의 RL 디컨벌루션은 계산에 많은 시간을 필요로 한다. 이를 개선하기 위하여 영상을 적절한 크기로 분할하여 각각 RL 디컨벌루션을 계산한다. 또한 분할 시 생기는 왜곡을 줄이기 위해 리플 제거를 위한 알고리듬을 추가한다. 이를 통해 기존의 알고리듬보다 연산 양을 줄여 빠른 RL 디컨벌루션이 가능하도록 개선한다.

• Proceedings of the IEEK Conference
• /
• 2007.07a
• /
• pp.279-280
• /
• 2007

Airway wall thickness is an important bio-marker for evaluation of pulmonary diseases such as stenosis, bronchiectasis. Nevertheless, an image-based analysis of the airway tree can provide precise and valuable airway size information, quantitative measurement of airway wall thickness in CT images involves various sources of error and uncertainty. So we have developed an accurate airway wall measurement technique for small airways with three-dimensional (3-D) approach. To illustrate performance of these techniques, we used airway phantom that consisted of 4 acryl tubes with various inner and outer diameters. Results show that evaluation of interpolation and deconvolution methods of airways in 3-D CT images, and significant improvement over the full-width-half-maximum method for measurement of not only location of the luminal and outer edge of the airway wall but airway wall thickness.

• The Korean Journal of Petroleum Geology
• /
• v.11 no.1 s.12
• /
• pp.9-17
• /
• 2005

In the main target area of the block II, Targe-scale faults occur below the unconformity developed around 1 km in depth. The contrast of seismic velocity around the unconformity is generally so large that the strong multiples and the radical velocity variation would deteriorate the quality of migrated section due to serious distortion. More than 15 kinds of data processing techniques have been applied to improve the image resolution for the structures farmed from this active crustal activity. The bad and noisy traces were edited on the common shot gathers in the first step to get rid of acquisition problems which could take place from unfavorable conditions such as climatic change during data acquisition. Correction of amplitude attenuation caused from spherical divergence and inelastic attenuation has been also applied. Mild F/K filter was used to attenuate coherent noise such as guided waves and side scatters. Predictive deconvolution has been applied before stacking to remove peg-leg multiples and water reverberations. The velocity analysis process was conducted at every 2 km interval to analyze migration velocity, and it was iterated to get the high fidelity image. The strum noise caused from streamer was completely removed by applying predictive deconvolution in time space and ${\tau}-P$ domain. Residual multiples caused from thin layer or water bottom were eliminated through parabolic radon transform demultiple process. The migration using curved ray Kirchhoff-style algorithm has been applied to stack data. The velocity obtained after several iteration approach for MVA (migration velocity analysis) was used instead or DMO for the migration velocity. Using various testing methods, optimum seismic processing parameter can be obtained for structural and stratigraphic interpretation in the Block II, Yellow Sea Basin.

• Proceedings of the IEEK Conference
• /
• 2001.09a
• /
• pp.547-550
• /
• 2001

Through matched filtering synthetic aperture radar (SAR) produces high-resolution imagery from data collected by a relative small antenna. While the impulse response obtained by the matched filter approach produces the best achievable signal-to-noise ratio, large sidelobes must be reduced to obtain higher-resolution SAR images. So, many enhancement methods of SAR imagery have been proposed. As a deconvolution method, the phase-extension inverse filtering is based on the characteristics of the matched filtering used in SAR imaging. It improves spatial resolution as well as effectively suppresses the sidelobes with low computational complexity. In the phase-extension inverse filtering, the impulse response is obtained from simulation with a point target. But in a real SAR environment, for example ERS-1, the impulse response is distorted by many non-ideal factors. So, in the phase-extension inverse filtering for a real SAR processing, the magnitudes of the frequency transfer function have to be compensated to produce more desirable results. In this paper, an estimation method to obtain a more accurate impulse response from a real SAR image is studied. And a compensation scheme to produce better performance of the phase-extension inverse filtering is also introduced.

• Journal of the Optical Society of Korea
• /
• v.19 no.2
• /
• pp.136-143
• /
• 2015

Image deblurring by a deconvolution method requires accurate knowledge of the blur kernel. Existing point spread function (PSF) models in the literature corresponding to lens aberrations and defocus are either parameterized and spatially invariant or spatially varying but discretely defined. In this paper, a parameterized model is developed and presented for a PSF which is spatially varying due to lens aberrations and defocus in an imaging system. The model is established from the Seidel third-order aberration coefficient and the Hu moment. A skew normal Gauss model is selected for parameterized PSF geometry structure. The accuracy of the model is demonstrated with simulations and measurements for a defocused infrared camera and a single spherical lens digital camera. Compared with optical software Code V, the visual results of two optical systems validate our analysis and proposed method in size, shape and direction. Quantitative evaluation results reveal the excellent accuracy of the blur kernel model.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.24 no.8B
• /
• pp.1558-1566
• /
• 1999

While high resolution images are required for various applications, aliased low-resolution images are only available due to the physical limitations of sensors. In this paper, we propose an algorithm to reconstruct a high resolution image from multiple aliased low-resolution images, which is based on the generalized multichannel deconvolution technique. The conventional approaches are based on the discrete Fourier transform (DFT) since the aliasing effect is easily analyzed in the frequency domain. However, the useful solution may not be available in many cases, i.e., the underdetermined cases or the insufficient subpixel information cases. In order to compensate for such ill-posedness, the generalized multichannel regularization was adopted in the spatial domain. Furthermore, the usage of the discrete cosine transform instead of the DFT leads to the computationally efficient reconstruction algorithm. The validity of the proposed algorithm is both theoretically and experimentally demonstrated in this paper. It is also shown that the effect of inaccurate motion information is reduced by regularization.

• Geophysics and Geophysical Exploration
• /
• v.12 no.2
• /
• pp.192-198
• /
• 2009

Side scan sonar and SBP (sub-bottom profiler) play a very important role in the survey for seafloor imaging and sub-bottom profiling. In this study, we have acquired side scan sonar and SBP data from the artificial reef area. We applied digital image processing techniques to side scan sonar data in order to improve an image quality. For the enhancement of data quality and image resolution, we applied the typical seismic data processing sequence including gain recovery, muting, spectrum analysis, predictive deconvolution, migration to SBP data. We could easily estimate if artificial reef structures were settled properly and their distribution on the seafloor from the integrated interpretation of side scan sonar and SBP data. From the sampling analysis of seabed sediments, texture filtering of side scan sonar data and SBP data interpretation, we could evaluate the sediment type, distribution and thickness of seafloor sediments in detail.

• Journal of Nuclear Fuel Cycle and Waste Technology(JNFCWT)
• /
• v.18 no.2
• /
• pp.217-225
• /
• 2020

Concrete waste accounts for approximately 70~80% of the total waste generated during the decommissioning of nuclear power plants (NPPs). Based upon the concentration of each radionuclide, the concrete waste from the decommissioning can be used in the determination of the clearance threshold used to classify waste as radioactive. To reduce the cost of radioactive concrete waste disposal, it is important to perform decontamination before self-disposal or limited recycling. Therefore, it is necessary to estimate the internal radioactivity distribution of radioactive concrete waste to ensure effective decontamination. In this study, the performance metrics of various Compton reconstruction algorithms were compared in order to identify the best strategy to estimate the internal radioactivity distribution in concrete waste during the decommissioning of NPPs. Four reconstruction algorithms, namely, simple back-projection, filtered back-projection, maximum likelihood expectation maximization (MLEM), and energy-deconvolution MLEM (E-MLEM) were used as Compton reconstruction algorithms. Subsequently, the results obtained by using these various reconstruction algorithms were compared with one another and evaluated, using quantitative evaluation methods. The MLEM and E-MLEM reconstruction algorithms exhibited the best performance in maintaining a high image resolution and signal-to-noise ratio (SNR), respectively. The results of this study demonstrate the feasibility of using Compton images in the estimation of the internal radioactive distribution of concrete during the decommissioning of NPPs.